Intelligent Multiagent Systems
Course Description
The course introduces a multiagent approach to distributed artificial intelligence. A notion of an inteligent rational agent as an inteligent system capable of autonomous, continuous and reactive action in an environment in pursuit of its goals. Taxonomy of agent architectures, formal languages for multiagent systems specification, languages and schemes for knowledge representation, formal languages and models for modeling of agent and environment behaviour, agent communication languages and associated semantic models. Basic interagent interaction patterns and coordination of cooperative and antagonistic agents. Coordination techniques: organisational structure, contracting, multiagent planning and negotiation. Application of multiagent systems in computer and robot vision, decision support systems, electronic commerce, robotics, and simulation of societies.
General Competencies
Providing an overview of the basic principles of multiagent paradigm. Acquainting students with formal approaches to multiagent system specification, knowledge representation, behaviour modeling and interagent communication in order to solve problems related to distributed artificial inteligence.
Learning Outcomes
- discuss the notions of the intelligent agent and multi-agent system
- distinguish basic categories of agents and multi-agent systems
- identify the basic application areas of intelligent agents and multi-agent systems
- apply basic multi-agent paradigms to the real world problem solving
- employ the basics of the game theory to formulate and solve multi-agent problems
- construct simple but functional multi-agent systems
Forms of Teaching
Lectures
Exams
Structural Exercises
Exams
Structural Exercises
Grading Method
Continuous Assessment | Exam | |||||
---|---|---|---|---|---|---|
Type | Threshold | Percent of Grade | Threshold | Percent of Grade | ||
Seminar/Project | 0 % | 20 % | 0 % | 20 % | ||
Mid Term Exam: Written | 0 % | 35 % | 0 % | |||
Final Exam: Written | 0 % | 45 % | ||||
Exam: Written | 0 % | 80 % |
Week by Week Schedule
- Distributed artificial intelligence. Multiagent approach. Multiagent systems and intelligent agents. Areas of application. Overview of related areas.
- Rational agents. Optimal decision making. Policy of the agent. Utility function. Markov decision process.
- Distributed constraint satisfaction.
- Introduction to the game theory. Payoff matrix. Solution strategies: maxmin strategy, social wellfare strategy, Pareto optimal strategy, iterative domination strategy, Nash equilibrium.
- Coordination and cooperation games. Characteristic form games and coalition formation.
- Learning in multi-agent systems.
- Representing knowledge of an inteligent agent. Inference. Nonmononotnic reasoning and belief revision. Knowledge and belief. Microtheories.
- Midterm exam
- Multiagent negotiation. Bargaining problem. Axiomatic solution concepts. Strategic solution concepts.
- Task allocation problem. Contracts. Complex deals. Argumentation-based negotiation. Negotiation networks. Network exchange theory.
- Auctions. Valuation function. Simple auctions: English auction, first-price sealed-bid auction, Dutch auction, Vickrey auction, double auction. Combinatorial auctions.
- Voting and mechanism design. Voting problem. Borda count. Groves-Clarke mechnanism. Vickrey-Groves-Clarke mechanism. Distributed mechanism design.
- Coordination using goal and plan hierarchies. TAEMS structure. Generalized partial global planning.
- Nature inspired approaches. Ants and termites. Immune systems.
- Final exam
Study Programmes
University graduate
[FER2-HR] Computer Science - profile
Specialization Course
(2. semester)
Literature
Lecturers
For students
General
ID 34545
Summer semester
4 ECTS
L1 English Level
L1 e-Learning
30 Lectures
0 Seminar
0 Exercises
0 Laboratory exercises
0 Project laboratory
Grading System
89 Excellent
76 Very Good
63 Good
50 Sufficient